Using Hierarchical Cluster Analysis to Generate Clusters of Agile Practices

Abstract

Agile software development methodologies consist of a set of software practices that can help organizations to produce products faster and deliver what customers want. Despite the benefits they gain from adopting agile practice, organizations could maximize the benefits gained by adopting correlated practices. There is a lack of study on the identification of clusters of independent practices. This paper focuses on identifying clusters of agile practices in software startups in Saudi Arabia. The study was conducted using a questionnaire with 76 software practitioners from software startups in Saudi Arabia. In this paper, 20 agile practices were analyzed using hierarchical cluster analysis. The analysis generated four clusters: Each was associated with a list of practices. These clusters can be used as a guide for agile method tailoring that helps to identify the inter-relationship between different agile practices. The clusters were labeled as project management, quality assurance, team communication, and incremental and iterative clusters. The results can be used to study the co-dependence of agile practices in depth. Moreover, the study can help adopters from similar domains as well as companies with limited resources and experience frequent changes in requirements to adopt these agile practice clusters.